Received: 20 April 2020 | Revised: 17 May 2020 | Accepted: 19 May 2020 DOI: 10.1002/jez.2388 RESEARCH PAPER Thermoregulation and thermal performance of crested geckos (Correlophus ciliatus) suggest an extended optimality hypothesis for the evolution of thermoregulatory set‐points Andrea Aparicio Ramirez | Karina Perez | Rory S. Telemeco Department of Biology, California State University Fresno, Fresno, California Abstract Correspondence Crested geckos (Correlophus ciliatus, formerly Rhacodactylus ciliatus) were redis- Rory S. Telemeco, Department of Biology, covered in New Caledonia 25 years ago and despite being common in the pet trade, California State University Fresno, Fresno, CA 93740. there is no published information on their physiology. We measured thermo- Email: [email protected] regulation (preferred body temperature, thermal set‐point range, and voluntary Funding information limits) and performance (thermal performance curves [TPC] for 25 cm sprint speed California State University Fresno and 1 m running speed) of adult and juvenile crested geckos in the laboratory to describe their thermal tolerances, differences among life stages, correlations be- tween behavior and performance, and correlations with natural temperatures. Despite lacking special lighting or heating requirements in captivity, crested geckos displayed typical thermal biology for a lizard with no difference among life stages. They thermoregulated to a narrow set‐point range (TSET,24–28°C), that broadly overlaps natural air temperatures in New Caledonia, during activity. Somewhat surprisingly, the optimal temperature for performance (TOPT, 32°C) was substantially above preferred body temperatures and approximated the average maximum tem- perature voluntarily experienced (VTMAX, 33°C). Preferred body temperatures, by contrast, corresponded to the lower threshold temperature (Td) where the TPC deviated from exponential, which we suggest is the temperature where performance is optimized after accounting for the costs of metabolic demand and overheating risk. Our results demonstrate that despite their lack of specific requirements when housed in human dwellings, crested geckos actively thermoregulate to temperatures that facilitate performance, and have thermal biology typical of other nocturnal or shade‐dwelling species. Additionally, crested geckos appear at little risk of direct climate change‐induced decline because increased temperatures should allow in- creased activity. KEYWORDS beta equation, lizard, New Caledonia, running performance, thermal performance curve, voluntary thermal limits (VTMAX and VTMIN) Andrea Aparicio Ramirez and Karina Perez contributed equally to this study. J Exp Zool. 2020;1–10. wileyonlinelibrary.com/journal/jez © 2020 Wiley Periodicals LLC | 1 2 | APARICIO RAMIREZ ET AL. 1 | INTRODUCTION populations, we must describe the average effects of temperature on performance as well as the variation, particularly among age classes. Because body temperature affects virtually all fitness‐linked pro- The TPC is predicted to evolve such that performance is cesses, thermal environment is an important selection pressure optimized at temperatures typically experienced by organisms that defines a major axis of the fundamental niche (reviewed in (Angilletta, 2009; Gilchrist, 1995). However, most animals do not Angilletta, 2009; McNab, 2002). Moreover, the thermal environment directly experience average environmental temperatures, but instead has proven highly dynamic in the Anthropocene, changing as a result alter their body temperatures via thermoregulation (Angilletta, 2009; of both direct habitat alteration and ongoing climate change Huey & Slatkin, 1976). Reptiles and other ectotherms primarily (IPCC, 2013; Lewis & Maslin, 2015; Steffen et al., 2018). Thus, a thermoregulate behaviorally by shuttling between thermal micro- necessary step for understanding both a species' evolutionary history environments (Cowles & Bogert, 1944; Huey & Slatkin, 1976). and prospect for long‐term persistence is to quantify the effects Because of their close link, the TPC and thermoregulatory behavior of temperature on behavior and performance (Bowler & are expected to coevolve, with thermoregulation under selection to Terblanche, 2008; Gunderson & Stillman, 2015; Taylor et al., 2020; optimize performance given the TPC, and the TPC under selection to Urban, Richardson, & Freidenfelds, 2014). This is especially true for optimize performance given body temperatures experienced via endemic, tropical species that are poorly studied, live in at‐risk ha- thermoregulation (reviewed in Angilletta, 2009). Selection on ther- bitat, and could have different thermal biology from widespread moregulation is expected to limit exposure to extreme or dangerous temperate species which tend to be better studied (although see temperatures thereby allowing persistence in diverse thermal en- substantial research on Caribbean Anolis for a tropical exception). vironments, but this reduces natural selection for increased thermal Here, we describe the thermal biology of such an endemic, the New tolerance (i.e., the Bogert effect; Huey, Hertz, & Sinervo, 2003; Caledonian crested gecko, Correlophus ciliatus (formerly Rhacodactylus Muñoz & Losos, 2018). As a result of this coevolution, individuals are ciliatus). expected to thermoregulate such that body temperatures closely The effect of body temperature on performance in ectotherms is match optimal temperatures during activity, so long as suitable classically described via the thermal performance curve (TPC; all thermal microenvironments are available and costs are minimal abbreviations are described in Table 1; Huey & Stevenson, 1979). In (Basson, Levy, Angilletta, & Clusella‐Trullas, 2017; Huey & reptiles, TPCs are typically left‐skewed and unimodal with whole‐ Slatkin, 1976; Sears et al., 2016). Optimal temperatures for ther- organism performance slowly increasing with body temperature moregulation could be those that maximize performance or those above the critical minimum (CTMIN) until an optimum temperature that maximize net performance after accounting for costs such as (TOPT) is reached where performance is maximized (Wf). Performance overheating risk (Angilletta, 2009; Huey & Bennett, 1987; Martin & then rapidly falls at temperatures above TOPT until reaching zero at Huey, 2008). Regardless, tight thermoregulation shields individuals the critical thermal maximum (CTMAX). Although the general shape of from natural selection for increased tolerance to extreme tempera- this relationship is well conserved, specific parameter values can vary tures, potentially resulting in narrow thermal tolerance breadth widely, both within and among species and even across the life span (Buckley & Huey, 2016; Muñoz et al., 2016; but see Logan, Cox, & of individuals (Careau, Biro, Bonneaud, Fokam, & Herrel, 2014; Calsbeek, 2014). Dowd, King, & Denny, 2015; Klockmann, Günter, & Fischer, 2016; Many environments, however, either do not allow or do not re- Rezende, Castañeda, & Santos, 2014; Sinclair et al., 2016). Thus, to quire such tight thermoregulation, potentially allowing increased understand how variation in the thermal environment affects selection on thermal tolerance. In the tropics, the temperature tends TABLE 1 Abbreviation Description Abbreviations and their descriptions CTMAX Critical thermal maximum. Here, CTMAX was estimated as the temperature where the TPC for running performance is predicted to drop to zero PBT Preferred body temperature Td Temperature of deviation from exponential. TPCs have both an upper and lower Td TOPT Optimum temperature where performance is maximized TPC Thermal performance curve TSET Thermal set‐point range VTMAX Voluntary thermal maximum VTMIN Voluntary thermal minimum Wf Peak performance. Occurs at the optimum temperature Note: Additional details are available in the main text. APARICIO RAMIREZ ET AL. | 3 to be relatively stable and warm throughout the year (Buckley & captive‐reared crested gecko and compared those with air tem- Huey, 2016; Pincebourde & Suppo, 2016; Sunday, Bates, & peratures in New Caledonia. We predicted that crested geckos would Dulvy, 2011). In such situations, animals might not need to actively not actively thermoregulate and maintain high performance across thermoregulate to maintain suitable body temperatures, and there temperatures commonly experienced. Furthermore, because juve- should be little selection for either tight behavioral thermoregulation niles can be harassed or even eaten by adults (de Vosjoli, 2005; or broad thermal tolerance (Buckley & Huey, 2016; Huey & Team, 2017), potentially forcing them to inhabit marginal habitats, Slatkin, 1976). By contrast, nocturnal species are expected to we predicted that juveniles would have broader thermal performance experience little thermal heterogeneity, but possibly cooler tem- breadth than adults. Finally, we directly compared thermal behavior peratures than diurnal species (Gunderson & Stillman, 2015). and performance to illuminate how natural selection has shaped Because thermal heterogeneity is required for behavioral thermo- thermal set‐points in this species. Based on this comparison, we regulation, this situation is expected to result in little selection for propose that natural selection shapes thermoregulatory set‐points to behavioral thermoregulation, but strong selection for performance at approximate the lower inflection point on the TPC (lower Td) because relatively cool temperatures. Additionally, life